The Real Reason Legacy B2B Is Doomed: Customers Are Becoming Micro-Vendors

When a $60K/year product breaks a core compliance feature like unsubscribe, you expect some urgency. But after two weeks of passing the blame — first to Salesforce, then to the customer — Jason Lemkin's team at SaaStr did what more and more businesses are doing: they built the fix themselves in an afternoon, using Replit and Claude.

Lemkin's right that this signals a cultural rot in legacy B2B. But as someone who tracks what builders actually do when their software fails, I think he's missing a deeper shift. It's not just that legacy vendors are lazy or arrogant — though many are. It's that the cost and speed of building your own solution have dropped so dramatically that customers are becoming micro-vendors of fixes for features their incumbent providers can't deliver.

At PainSignal, we track problems users post about software, and their ideas for apps that solve them. Here's the number that should keep legacy B2B leaders awake: we've logged 128 app ideas where users describe building custom tools to replace a core feature of a legacy platform. The median cost estimate these builders cite? Under $1,000. Compare that to Marketo's $60K annual bill.

These aren't pie-in-the-sky pitches on a whiteboard. They're people who already shipped a replacement for something their current vendor can't get right. Unsubscribe handling. Report generation. Data export. Workflow automation. The list reads like a greatest hits of B2B software's weakest links.

The support collapse is a leading indicator

Lemkin's Marketo support saga is excruciatingly familiar. Our data shows 78% of support-related problems tagged on PainSignal involve legacy B2B software, with "no accountability" as the top complaint. We track 47 specific problems in legacy marketing automation with an average severity of 3.8 out of 5. Broken unsubscribes, delayed campaign execution, and data sync failures dominate.

When engineering blames the customer for a CAN-SPAM violation, it's not just a bad support interaction. It tells every technical buyer in earshot: "You should build your own fix, because we won't." Five years ago, that meant expensive custom development. Today, it means lunchtime in Replit.

Our data also shows 32 problems with poor API documentation and rate limits in marketing platforms, with users reporting difficulty connecting AI agents to these systems. Lemkin noted Marketo's API bottleneck — one good read per day. That's not just annoying; it's disqualifying when every modern tool you compare against (Stripe, Supabase, Linear) treats API design as a first-class product.

The macro multiplier

Lemkin argues B2B stocks are crashing mainly because product quality has fallen behind the AI age. I agree product quality is a necessary condition, but our data suggests it's not sufficient on its own. Of the vendor switching decisions we track, 60% cite pricing increases and contract lock-ins as equal or greater triggers. The macroeconomic pressure of 2025-2026 amplifies impatience. When a vendor hikes renewal 30% and delivers a broken unsubscribe link, it's not just one failure — it's a gut-check moment.

That's why I think the "build your own" trend is the hidden accelerant. It doesn't require a formal RFP or a sales process. One afternoon, one frustrated developer, one Claude session. The customer doesn't leave the vendor entirely — yet. But they've just surgically removed the vendor from a core workflow. They've become a partial vendor competitor, and their next project will be even bigger.

What this means for builders and investors

For indie hackers and vibe coders: there's a goldmine in specific feature replacements that legacy vendors consistently fail at. The 128 app ideas we track span 15+ categories, from compliance-critical functions like unsubscribe handling to more routine operations like lead scoring and data normalization. Low acquisition costs because the user already knows the pain. No need to build a whole CRM — just the part that Marketo won't fix.

For seed investors: track support-quality signals as a precursor to churn. A vendor that blames customers for bugs, as Lemkin documents, isn't just losing that one renewal. They're training an entire generation of technical buyers to distrust them — and to build their own escape hatches.

The software isn't just not good enough for the AI age. It's not good enough compared to what a team of three with Claude can ship in a day. That's a much lower bar, and legacy B2B is clearing it less and less often.